Data visualisation in R using ggplot2
if (!require("pacman")) install.packages("pacman")
pacman::p_load(
tidyverse,
data.table, # fast csv read/write
janitor, # clean_names()
scales, # comma()
Cairo, # better anti-aliasing when saving images
ggtext, # for markdown text
ggdist, # for many distribution geoms
palmerpenguins # example dataset on penguins
)Below, we’ll visualise the palmer penguins dataset to introduce the basics of ggplot2.
penguinsFirstly, we use ggplot2 to plot a basic scatter plot of flipper length against bill length.
We define 2 aesthetics with aes() as the x and y coordinates, then one geometric object with geom_point() to plot these coordinates as points.
Extra things to note:
The x and y coordinates can also be specified by using x = and y = but this is not required because ggplot2 assumes that the first two arguments after data will be x and y aesthetics, respectively.
Awkwardly, the + pipe is used after the ggplot() function, as the %>% pipe was developed after ggplot2.
penguins %>%
ggplot(aes(flipper_length_mm, bill_length_mm)) +
geom_point()
We can now define colour by including it as another aesthetic inside aes().
penguins %>%
ggplot(aes(flipper_length_mm, bill_length_mm,
colour = species)) +
geom_point()
The same can be done with the size aesthetic.
penguins %>%
ggplot(aes(flipper_length_mm, bill_length_mm,
colour = species,
size = body_mass_g)) +
geom_point()
The alpha (transparency) of the points can also be set as an aesthetic, or just manually to 50% as we do here.
penguins %>%
ggplot(aes(flipper_length_mm, bill_length_mm,
colour = species,
size = body_mass_g)) +
geom_point(alpha = 0.5)
The plot can be split by a categorical/factor variable into facets using facet_wrap().
penguins %>%
ggplot(aes(flipper_length_mm, bill_length_mm,
colour = species,
size = body_mass_g)) +
geom_point(alpha = 0.5) +
facet_wrap(~island)
This plot can be saved using the ggsave() function.
ggsave(dpi = 300, width = 7, height = 5, units = "in", type = "cairo",
filename = "outputs/penguins5.jpeg")
Example using ggplot2 with covid vaccination data
Example MSOA level covid data - https://coronavirus.data.gov.uk/details/download
vaccines <- fread("https://api.coronavirus.data.gov.uk/v2/data?areaType=msoa&metric=cumVaccinationThirdInjectionUptakeByVaccinationDatePercentage&format=csv") %>%
clean_names() %>%
rename(uptake = cum_vaccination_third_injection_uptake_by_vaccination_date_percentage) %>%
select(region_name,
area_name,
uptake)Plot the % vaccine uptake for each MSOA, split across the regions of England using geom_point().
dist_1 <- vaccines %>%
ggplot(aes(uptake, region_name)) +
geom_point()
ggsave(dist_1, dpi = 300, width = 12, height = 12, units = "in", type = "cairo",
filename = "outputs/dist_1.jpeg")
Replace geom_point() with geom_jitter() to randomly “jitter” the points and reduce overplotting.
dist_2 <- vaccines %>%
ggplot(aes(uptake, region_name)) +
geom_jitter()
ggsave(dist_2, dpi = 300, width = 12, height = 12, units = "in", type = "cairo",
filename = "outputs/dist_2.jpeg")
Set point transparency to 40% to reduce overplotting further.
dist_3 <- vaccines %>%
ggplot(aes(uptake, region_name)) +
geom_jitter(alpha = 0.4)
ggsave(dist_3, dpi = 300, width = 12, height = 12, units = "in", type = "cairo",
filename = "outputs/dist_3.jpeg")
Add boxplots with geom_boxplot().
dist_4 <- vaccines %>%
ggplot(aes(uptake, region_name)) +
geom_jitter(alpha = 0.4) +
geom_boxplot()
ggsave(dist_4, dpi = 300, width = 12, height = 12, units = "in", type = "cairo",
filename = "outputs/dist_4.jpeg")
Set boxplot transparency and remove duplicate outliers.
dist_5 <- vaccines %>%
ggplot(aes(uptake, region_name)) +
geom_jitter(alpha = 0.4) +
geom_boxplot(alpha = 0.7, outlier.shape = NA)
ggsave(dist_5, dpi = 300, width = 12, height = 12, units = "in", type = "cairo",
filename = "outputs/dist_5.jpeg")
Increase text size with theme().
dist_6 <- vaccines %>%
ggplot(aes(uptake, region_name)) +
geom_jitter(alpha = 0.4) +
geom_boxplot(alpha = 0.7, outlier.shape = NA) +
theme(text = element_text(size = 16))
ggsave(dist_6, dpi = 300, width = 12, height = 12, units = "in", type = "cairo",
filename = "outputs/dist_6.jpeg")
Order by median % uptake.
dist_7 <- vaccines %>%
group_by(region_name) %>%
mutate(uptake_m = median(uptake, na.rm = TRUE)) %>%
ungroup() %>%
ggplot(aes(uptake, fct_reorder(region_name, uptake_m))) +
geom_jitter(alpha = 0.4) +
geom_boxplot(alpha = 0.7, outlier.shape = NA) +
theme(text = element_text(size = 16))
ggsave(dist_7, dpi = 300, width = 12, height = 12, units = "in", type = "cairo",
filename = "outputs/dist_7.jpeg")
Change some colours to suit an DHSC style and declutter by altering theme() to remove extra gridlines, remove axis ticks, and change the background to white.
dist_8 <- vaccines %>%
group_by(region_name) %>%
mutate(uptake_m = median(uptake, na.rm = TRUE)) %>%
ungroup() %>%
ggplot(aes(uptake, fct_reorder(region_name, uptake_m))) +
geom_jitter(colour = "#061C23", alpha = 0.4) +
geom_boxplot(colour = "#061C23", alpha = 0.7, outlier.shape = NA) +
theme(panel.background = element_rect(fill = "white", colour = NA),
plot.background = element_rect(fill = "white", colour = NA),
text = element_text(size = 16),
panel.grid.major.x = element_line(colour = "grey95"),
axis.ticks = element_blank())
ggsave(dist_8, dpi = 300, width = 12, height = 12, units = "in", type = "cairo",
filename = "outputs/dist_8.jpeg")
Add a title and correct axis text with labs().
dist_9 <- vaccines %>%
mutate(region_name = case_when(
region_name == "Yorkshire and The Humber" ~
"Yorkshire and\nThe Humber",
TRUE ~ as.character(region_name))) %>%
group_by(region_name) %>%
mutate(uptake_m = median(uptake, na.rm = TRUE)) %>%
ungroup() %>%
ggplot(aes(uptake, fct_reorder(region_name, uptake_m))) +
geom_jitter(colour = "#061C23", alpha = 0.4) +
geom_boxplot(colour = "#061C23", alpha = 0.7, outlier.shape = NA) +
labs(x = "Third vaccination uptake (%)",
y = NULL,
title = "Total uptake of third covid-19 vaccines vary across England") +
theme(panel.background = element_rect(fill = "white", colour = NA),
plot.background = element_rect(fill = "white", colour = NA),
text = element_text(size = 16),
panel.grid.major.x = element_line(colour = "grey95"),
axis.ticks = element_blank())
ggsave(dist_9, dpi = 300, width = 12, height = 12, units = "in", type = "cairo",
filename = "outputs/dist_9.jpeg")
Alter axis and margin spacings with scale_x_continuous(), scale_y_discrete(), and theme().
dist_10 <- vaccines %>%
mutate(region_name = case_when(
region_name == "Yorkshire and The Humber" ~
"Yorkshire and\nThe Humber",
TRUE ~ as.character(region_name))) %>%
group_by(region_name) %>%
mutate(uptake_m = median(uptake, na.rm = TRUE)) %>%
ungroup() %>%
ggplot(aes(uptake, fct_reorder(region_name, uptake_m))) +
geom_jitter(colour = "#061C23", alpha = 0.4) +
geom_boxplot(colour = "#061C23", alpha = 0.7, outlier.shape = NA) +
scale_x_continuous(expand = c(0.01, 0.01)) +
scale_y_discrete(expand = c(0.05, 0)) +
labs(x = "Third vaccination uptake (%)",
y = NULL,
title = "Total uptake of third covid-19 vaccines vary across England") +
theme(panel.background = element_rect(fill = "white", colour = NA),
plot.background = element_rect(fill = "white", colour = NA),
axis.ticks = element_blank(),
panel.grid.major.x = element_line(colour = "grey95"),
text = element_text(size = 16),
plot.title.position = "plot",
plot.margin = margin(10, 10, 10, 10))
ggsave(dist_10, dpi = 300, width = 12, height = 12, units = "in", type = "cairo",
filename = "outputs/dist_10.jpeg")
Increase text size further, change font, and alter colours with theme().
dist_11 <- vaccines %>%
mutate(region_name = case_when(
region_name == "Yorkshire and The Humber" ~
"Yorkshire and\nThe Humber",
TRUE ~ as.character(region_name))) %>%
group_by(region_name) %>%
mutate(uptake_m = median(uptake, na.rm = TRUE)) %>%
ungroup() %>%
ggplot(aes(uptake, fct_reorder(region_name, uptake_m))) +
geom_jitter(colour = "#061C23", alpha = 0.4) +
geom_boxplot(colour = "#061C23", alpha = 0.7, outlier.shape = NA) +
scale_x_continuous(expand = c(0.01, 0.01)) +
scale_y_discrete(expand = c(0.05, 0)) +
labs(x = "Third vaccination uptake (%)",
y = NULL,
title = "Total uptake of third covid-19 vaccines vary across England") +
theme(panel.background = element_rect(fill = "white", colour = NA),
plot.background = element_rect(fill = "white", colour = NA),
axis.ticks = element_blank(),
panel.grid.major.x = element_line(colour = "grey95"),
text = element_text(family = "sans", colour = "#061C23",
face = "bold", size = 16),
axis.text.y = element_text(family = "sans", colour = "#061C23",
face = "bold", size = 16),
plot.title.position = "plot",
plot.title = element_text(family = "sans", colour = "#061C23",
face = "bold", size = 29),
plot.margin = margin(10, 10, 10, 10))
ggsave(dist_11, dpi = 300, width = 12, height = 12, units = "in", type = "cairo",
filename = "outputs/dist_11.jpeg")
Replace the points and boxplots with stat_dots() from the ggdist package.
dist_12 <- vaccines %>%
mutate(region_name = case_when(
region_name == "Yorkshire and The Humber" ~
"Yorkshire and\nThe Humber",
TRUE ~ as.character(region_name))) %>%
group_by(region_name) %>%
mutate(uptake_m = median(uptake, na.rm = TRUE)) %>%
ungroup() %>%
ggplot(aes(uptake, fct_reorder(region_name, uptake_m))) +
stat_dots(colour = "#061C23", side = "top", scale = 0.74, shape = 15) +
scale_x_continuous(expand = c(0.01, 0.01)) +
scale_y_discrete(expand = c(0.05, 0)) +
labs(x = "Third vaccination uptake (%)",
y = NULL,
title = "Total uptake of third covid-19 vaccines vary across England") +
theme(panel.background = element_rect(fill = "white", colour = NA),
plot.background = element_rect(fill = "white", colour = NA),
axis.ticks = element_blank(),
panel.grid.major.x = element_line(colour = "grey95"),
text = element_text(family = "sans", colour = "#061C23",
face = "bold", size = 16),
axis.text.y = element_text(family = "sans", colour = "#061C23",
face = "bold", size = 16),
plot.title.position = "plot",
plot.title = element_text(family = "sans", colour = "#061C23",
face = "bold", size = 29),
plot.margin = margin(10, 10, 10, 10))
ggsave(dist_12, dpi = 300, width = 12, height = 12, units = "in", type = "cairo",
filename = "outputs/dist_12.jpeg")
Include stat_interval() to indicate where 25%, 50%, 95%, and 100% of the MSOAs lie.
dist_13 <- vaccines %>%
mutate(region_name = case_when(
region_name == "Yorkshire and The Humber" ~
"Yorkshire and\nThe Humber",
TRUE ~ as.character(region_name))) %>%
group_by(region_name) %>%
mutate(uptake_m = median(uptake, na.rm = TRUE)) %>%
ungroup() %>%
ggplot(aes(uptake, fct_reorder(region_name, uptake_m))) +
stat_dots(colour = "#061C23", side = "top", scale = 0.74, shape = 15) +
stat_interval(aes(color_ramp = stat(level)),
.width = c(0.25, 0.5, 0.95, 1),
position = position_dodge(width = 0.15, preserve = "single")) +
scale_x_continuous(expand = c(0.01, 0.01)) +
scale_y_discrete(expand = c(0.05, 0)) +
scale_colour_manual(values = c("grey78", "#79BDD3", "#4CA7C4", "#2092B6")) +
labs(x = "Third vaccination uptake (%)",
y = NULL,
title = "Total uptake of third covid-19 vaccines vary across England") +
theme(panel.background = element_rect(fill = "white", colour = NA),
plot.background = element_rect(fill = "white", colour = NA),
axis.ticks = element_blank(),
panel.grid.major.x = element_line(colour = "grey95"),
text = element_text(family = "sans", colour = "#061C23",
face = "bold", size = 16),
axis.text.y = element_text(family = "sans", colour = "#061C23",
face = "bold", size = 16),
plot.title.position = "plot",
plot.title = element_text(family = "sans", colour = "#061C23",
face = "bold", size = 29),
plot.margin = margin(10, 10, 10, 10))
ggsave(dist_13, dpi = 300, width = 12, height = 12, units = "in", type = "cairo",
filename = "outputs/dist_13.jpeg")
Declutter by replacing the legend with a colour coded subtitle, using the ggtext package.
dist_14 <- vaccines %>%
mutate(region_name = case_when(
region_name == "Yorkshire and The Humber" ~
"Yorkshire and\nThe Humber",
TRUE ~ as.character(region_name))) %>%
group_by(region_name) %>%
mutate(uptake_m = median(uptake, na.rm = TRUE)) %>%
ungroup() %>%
ggplot(aes(uptake, fct_reorder(region_name, uptake_m))) +
stat_dots(colour = "#061C23", side = "top", scale = 0.74, shape = 15) +
stat_interval(aes(color_ramp = stat(level)),
.width = c(0.25, 0.5, 0.95, 1),
position = position_dodge(width = 0.15, preserve = "single")) +
scale_x_continuous(expand = c(0.01, 0.01)) +
scale_y_discrete(expand = c(0.05, 0)) +
scale_colour_manual(values = c("grey78", "#79BDD3", "#4CA7C4", "#2092B6")) +
labs(x = "Third vaccination uptake (%)",
y = NULL,
title = "Total uptake of third covid-19 vaccines vary across England",
subtitle = "The triangles indicate the median vaccine uptake of MSOAs in
each region. Each square represents one MSOA. The interval stripes
indicate the ranges in which
**<span style='color:#2092B6;'>25%</span>**,
**<span style='color:#4CA7C4;'>50%</span>**,
**<span style='color:#79BDD3;'>95%</span>**, and
**<span style='color:grey78;'>100%</span>** of the MSOAs fall.") +
theme(legend.position = "none",
panel.background = element_rect(fill = "white", colour = NA),
plot.background = element_rect(fill = "white", colour = NA),
axis.ticks = element_blank(),
panel.grid.major.x = element_line(colour = "grey95"),
text = element_text(family = "sans", colour = "#061C23",
face = "bold", size = 16),
axis.text.y = element_text(family = "sans", colour = "#061C23",
face = "bold", size = 16),
plot.title.position = "plot",
plot.title = element_text(family = "sans", colour = "#061C23",
face = "bold", size = 29),
plot.subtitle = element_textbox_simple(family = "sans", face = "bold",
size = 16,
margin = margin(10, 0, -15, 0)),
plot.margin = margin(10, 10, 10, 10))
ggsave(dist_14, dpi = 300, width = 12, height = 12, units = "in", type = "cairo",
filename = "outputs/dist_14.jpeg")
Add geom_point() to plot the median values only.
dist_15 <- vaccines %>%
mutate(region_name = case_when(
region_name == "Yorkshire and The Humber" ~
"Yorkshire and\nThe Humber",
TRUE ~ as.character(region_name))) %>%
group_by(region_name) %>%
mutate(uptake_m = median(uptake, na.rm = TRUE)) %>%
ungroup() %>%
ggplot(aes(uptake, fct_reorder(region_name, uptake_m))) +
stat_dots(colour = "#061C23", side = "top", scale = 0.74, shape = 15) +
stat_interval(aes(color_ramp = stat(level)),
.width = c(0.25, 0.5, 0.95, 1),
position = position_dodge(width = 0.15, preserve = "single")) +
geom_point(aes(uptake_m, fct_reorder(region_name, uptake_m)),
size = 5, colour = "#061C23") +
scale_x_continuous(expand = c(0.01, 0.01)) +
scale_y_discrete(expand = c(0.05, 0)) +
scale_colour_manual(values = c("grey78", "#79BDD3", "#4CA7C4", "#2092B6")) +
labs(x = "Third vaccination uptake (%)",
y = NULL,
title = "Total uptake of third covid-19 vaccines vary across England",
subtitle = "The triangles indicate the median vaccine uptake of MSOAs in
each region. Each square represents one MSOA. The interval stripes
indicate the ranges in which
**<span style='color:#2092B6;'>25%</span>**,
**<span style='color:#4CA7C4;'>50%</span>**,
**<span style='color:#79BDD3;'>95%</span>**, and
**<span style='color:grey78;'>100%</span>** of the MSOAs fall.") +
theme(legend.position = "none",
panel.background = element_rect(fill = "white", colour = NA),
plot.background = element_rect(fill = "white", colour = NA),
axis.ticks = element_blank(),
panel.grid.major.x = element_line(colour = "grey95"),
text = element_text(family = "sans", colour = "#061C23",
face = "bold", size = 16),
axis.text.y = element_text(family = "sans", colour = "#061C23",
face = "bold", size = 16),
plot.title.position = "plot",
plot.title = element_text(family = "sans", colour = "#061C23",
face = "bold", size = 29),
plot.subtitle = element_textbox_simple(family = "sans", face = "bold",
size = 16,
margin = margin(10, 0, -15, 0)),
plot.margin = margin(10, 10, 10, 10))
ggsave(dist_15, dpi = 300, width = 12, height = 12, units = "in", type = "cairo",
filename = "outputs/dist_15.jpeg")
Alter the median point shape and position.
dist_16 <- vaccines %>%
mutate(region_name = case_when(
region_name == "Yorkshire and The Humber" ~
"Yorkshire and\nThe Humber",
TRUE ~ as.character(region_name))) %>%
group_by(region_name) %>%
mutate(uptake_m = median(uptake, na.rm = TRUE)) %>%
ungroup() %>%
ggplot(aes(uptake, fct_reorder(region_name, uptake_m))) +
stat_dots(colour = "#061C23", side = "top", scale = 0.74, shape = 15) +
stat_interval(aes(color_ramp = stat(level)),
.width = c(0.25, 0.5, 0.95, 1),
position = position_dodge(width = 0.15, preserve = "single")) +
geom_point(aes(uptake_m, fct_reorder(region_name, uptake_m)),
shape = 17, size = 5, colour = "#061C23",
position = position_dodge(width = 0.42, preserve = "single")) +
scale_x_continuous(expand = c(0.01, 0.01)) +
scale_y_discrete(expand = c(0.05, 0)) +
scale_colour_manual(values = c("grey78", "#79BDD3", "#4CA7C4", "#2092B6")) +
labs(x = "Third vaccination uptake (%)",
y = NULL,
title = "Total uptake of third covid-19 vaccines vary across England",
subtitle = "The triangles indicate the median vaccine uptake of MSOAs in
each region. Each square represents one MSOA. The interval stripes
indicate the ranges in which
**<span style='color:#2092B6;'>25%</span>**,
**<span style='color:#4CA7C4;'>50%</span>**,
**<span style='color:#79BDD3;'>95%</span>**, and
**<span style='color:grey78;'>100%</span>** of the MSOAs fall.") +
theme(legend.position = "none",
panel.background = element_rect(fill = "white", colour = NA),
plot.background = element_rect(fill = "white", colour = NA),
axis.ticks = element_blank(),
panel.grid.major.x = element_line(colour = "grey95"),
text = element_text(family = "sans", colour = "#061C23",
face = "bold", size = 16),
axis.text.y = element_text(family = "sans", colour = "#061C23",
face = "bold", size = 16),
plot.title.position = "plot",
plot.title = element_text(family = "sans", colour = "#061C23",
face = "bold", size = 29),
plot.subtitle = element_textbox_simple(family = "sans", face = "bold",
size = 16,
margin = margin(10, 0, -15, 0)),
plot.margin = margin(10, 10, 10, 10))
ggsave(dist_16, dpi = 300, width = 12, height = 12, units = "in", type = "cairo",
filename = "outputs/dist_16.jpeg")
Add geom_text() labels to include the counts of MSOAs.
dist_17 <- vaccines %>%
mutate(region_name = case_when(
region_name == "Yorkshire and The Humber" ~
"Yorkshire and\nThe Humber",
TRUE ~ as.character(region_name))) %>%
group_by(region_name) %>%
mutate(uptake_m = median(uptake, na.rm = TRUE),
count = n()) %>%
ungroup() %>%
ggplot(aes(uptake, fct_reorder(region_name, uptake_m))) +
stat_dots(colour = "#061C23", side = "top", scale = 0.74, shape = 15) +
stat_interval(aes(color_ramp = stat(level)),
.width = c(0.25, 0.5, 0.95, 1),
position = position_dodge(width = 0.15, preserve = "single")) +
geom_text(aes(uptake_m,
as.numeric(fct_reorder(region_name, uptake_m)) - 0.185,
label = paste0(
" ",
comma(round_half_up(uptake_m, 0), accuracy = 1),
"% (", comma(count, accuracy = 1), " MSOAs)"
)),
fontface = "bold", size = 12/.pt, colour = "#061C23",
hjust = 0, check_overlap = TRUE) +
geom_point(aes(uptake_m, fct_reorder(region_name, uptake_m)),
shape = 17, size = 5, colour = "#061C23",
position = position_dodge(width = 0.42, preserve = "single")) +
scale_x_continuous(expand = c(0.01, 0.01)) +
scale_y_discrete(expand = c(0.05, 0)) +
scale_colour_manual(values = c("grey78", "#79BDD3", "#4CA7C4", "#2092B6")) +
labs(x = "Third vaccination uptake (%)",
y = NULL,
title = "Total uptake of third covid-19 vaccines vary across England",
subtitle = "The triangles indicate the median vaccine uptake of MSOAs in
each region. Each square represents one MSOA. The interval stripes
indicate the ranges in which
**<span style='color:#2092B6;'>25%</span>**,
**<span style='color:#4CA7C4;'>50%</span>**,
**<span style='color:#79BDD3;'>95%</span>**, and
**<span style='color:grey78;'>100%</span>** of the MSOAs fall.") +
theme(legend.position = "none",
panel.background = element_rect(fill = "white", colour = NA),
plot.background = element_rect(fill = "white", colour = NA),
axis.ticks = element_blank(),
panel.grid.major.x = element_line(colour = "grey95"),
text = element_text(family = "sans", colour = "#061C23",
face = "bold", size = 16),
axis.text.y = element_text(family = "sans", colour = "#061C23",
face = "bold", size = 16),
plot.title.position = "plot",
plot.title = element_text(family = "sans", colour = "#061C23",
face = "bold", size = 29),
plot.subtitle = element_textbox_simple(family = "sans", face = "bold",
size = 16,
margin = margin(10, 0, -15, 0)),
plot.margin = margin(10, 10, 10, 10))
ggsave(dist_17, dpi = 300, width = 12, height = 12, units = "in", type = "cairo",
filename = "outputs/dist_17.jpeg")
Declutter by removing the x axis entirely. Was this a good idea?
dist18 <- vaccines %>%
mutate(region_name = case_when(
region_name == "Yorkshire and The Humber" ~
"Yorkshire and\nThe Humber",
TRUE ~ as.character(region_name))) %>%
group_by(region_name) %>%
mutate(uptake_m = median(uptake, na.rm = TRUE),
count = n()) %>%
ungroup() %>%
ggplot(aes(uptake, fct_reorder(region_name, uptake_m))) +
stat_dots(colour = "#061C23", side = "top", scale = 0.74, shape = 15) +
stat_interval(aes(color_ramp = stat(level)),
.width = c(0.25, 0.5, 0.95, 1),
position = position_dodge(width = 0.15, preserve = "single")) +
geom_text(aes(uptake_m,
as.numeric(fct_reorder(region_name, uptake_m)) - 0.185,
label = paste0(
" ",
comma(round_half_up(uptake_m, 0), accuracy = 1),
"% (", comma(count, accuracy = 1), " MSOAs)"
)),
fontface = "bold", size = 12/.pt, colour = "#061C23",
hjust = 0, check_overlap = TRUE) +
geom_point(aes(uptake_m, fct_reorder(region_name, uptake_m)),
shape = 17, size = 5, colour = "#061C23",
position = position_dodge(width = 0.42, preserve = "single")) +
scale_x_continuous(expand = c(0.01, 0.01)) +
scale_y_discrete(expand = c(0.05, 0)) +
scale_colour_manual(values = c("grey78", "#79BDD3", "#4CA7C4", "#2092B6")) +
labs(x = NULL,
y = NULL,
title = "Total uptake of third covid-19 vaccines vary across England",
subtitle = "The triangles indicate the median vaccine uptake of MSOAs in
each region. Each square represents one MSOA. The interval stripes
indicate the ranges in which
**<span style='color:#2092B6;'>25%</span>**,
**<span style='color:#4CA7C4;'>50%</span>**,
**<span style='color:#79BDD3;'>95%</span>**, and
**<span style='color:grey78;'>100%</span>** of the MSOAs fall.") +
theme(legend.position = "none",
panel.background = element_rect(fill = "white", colour = NA),
plot.background = element_rect(fill = "white", colour = NA),
axis.text.x = element_blank(),
axis.ticks = element_blank(),
axis.text.y = element_text(family = "sans", colour = "#061C23",
face = "bold", size = 16),
plot.title.position = "plot",
plot.title = element_text(family = "sans", colour = "#061C23",
face = "bold", size = 29),
plot.subtitle = element_textbox_simple(family = "sans", face = "bold",
size = 16,
margin = margin(10, 0, -15, 0)),
plot.margin = margin(10, 10, 10, 10))
ggsave(dist18, dpi = 300, width = 12, height = 12, units = "in", type = "cairo",
filename = "outputs/dist18.jpeg")